Over the past few months, I've become increasingly convinced that we're on the cusp of a transformative cycle that will rival the rise of the Internet. And it's going to hit users where they live: the desktops, laptops, and mobile devices they use every day.
At the heart of the matter is virtualization. In the data center, no technology I can think of has enjoyed the hockey-stick adoption curve exhibited by server virtualization, where admins can now scale capacity by spinning up virtual machines with the click of a mouse, instead of provisioning physical hardware.
[ Also on InfoWorld: Find out "What desktop virtualization really means." | Read Neil McAllister's eyewitness account of Google's Chrome OS announcement. | Download Paul Venezia's comprehensive Thin Client Computing Deep Dive Report. ]
Yet for all the scalability, availability, and hardware utilization benefits of server virtualization, I believe there's an even stronger argument for wrapping up user desktops as virtual machines (although, as you will see, I mean that in an abstract sense). There's more to it than the usual pitch for desktop virtualization -- which asserts that desktops should be packaged as VMs and run on a server because physical desktops now incur unbearable security risks and administrative overhead. That's only half the story.
The other half is portability. The user environment people experience today as tied to a physical desktop device should not simply slink into the data center. That virtual machine should be fully portable, so users can take it with them anywhere, even if they lack a network connection; also, when they reconnect, it should sync with the server VM. Plus, that virtual machine should be downloadable -- or, worst case, remotely accessible -- from any device. Ultimately, the operating system UI for that VM should adapt itself to whatever physical device is available, from a PC in an Internet café to an iPad.
That day may seem far away, but already, two approaches to the fully portable VM have worked up a head of steam. First and most obvious is the client hypervisor, released in "test kit" form by Citrix Systems and currently under development by EMC VMware. If client hypervisors reach their full potential, users should be able to carry their VMs with them on a wide range of laptops or other mobile hardware, compute without a connection, and sync with the server when that connection is restored.
The second approach is the one being pursued by Google with its Chrome OS. As Google has characterized it, Chrome OS is merely a Linux kernel with the Chrome browser as its shell -- basically an operating system for a cheap Web appliance. Ultimately it will be more than that.
Already, Google Gears lets developers create Web applications that run without a connection. You can be sure Google is also working on the cloud equivalent of a virtual machine for users -- a personal, customizable desktop environment, complete with applications and preferences, accessible from any device running a Chrome browser. Google's recent $30 million purchase of Bump Technologies, whose wildly imaginative BumpTop UI turned heads at the TED conference one year ago, is just the latest indication that Google may be moving in this direction.
So pay no attention to the idea that the surge in mobile computing will marginalize the desktop or that desktop virtualization will send us back in time to generic, mainframe-style computing. The new, virtual desktop may technically reside in your company's datacenter or on a server in the cloud, but that doesn't mean it's slipping away from you. In fact, not so many years from now, that reliable old companion in its new virtual form will follow you wherever you go -- and you'll be able to interact with it through whatever computing device is near at hand.
This article, "The coming rebirth of the desktop PC" originally appeared at InfoWorld.com. Read more of Eric Knorr's Modernizing IT blog and get a digest of the key stories each day in the InfoWorld Daily newsletter and on your mobile device at infoworldmobile.com.